b i o m a s s a n d b i o e n e r g y x x x ( 2 0 1 4 ) 1 e1 2

Available online at www.sciencedirect.com

ScienceDirect
http://www.elsevier.com/locate/biombioe

Aboveground and belowground biomass allocation

in native Prosopis caldenia Burkart secondaries
woodlands in the semi-arid Argentinean pampas
Lucia Risio a,b,*, Celia Herrero a,b, Stella M. Bogino c, Felipe Bravo a,b
a

Sustainable Forest Management Research Institute University of Valladolid-INIA, Av. Madrid 44, 34004 Palencia,
Spain
b
Departamento de Produccion Vegetal y Recursos Forestales, E.T.S. de Ingenieras Agrarias, Universidad de
Valladolid, Campus de Palencia, Spain
c
Departamento de Ciencias Agropecuarias, Facultad de Ingeniera y Ciencias Agropecuarias, Universidad Nacional
de San Luis, Argentina

article info

abstract

Article history:

The woodlands in the south-west of the Argentinean pampas are dominated by Prosopis

Received 29 October 2013

Caldenia Burkart (calden). The current deforestation rate of this woodlands is 0.82% per

Received in revised form

year. Different compensation initiatives have begun that recognize the role of forests as

14 March 2014

environmental service providers. The financial incentives they offer make it necessary to

Accepted 17 March 2014

quantify the amount of carbon stored in the forest biomass. A model for estimating calden

Available online xxx

biomass was developed. Thirty-eight trees were selected, felled and divided into sections.
An equation system was fitted using joint generalized regression to ensure the additivity

Keywords:

property. A weighted regression was used to avoid heteroscedasticity. In these woodlands

Calden

fire is the main disturbance and it can modify tree allometry, due this all models included

Aboveground biomass

the area of the base of the stem and tree height as independent variables since it indirectly

Belowground biomass

collects this variability. Total biomass and the stem fraction had the highest R2_Adj. values

Additivity

(0.75), while branches with a diameter less than 7 cm had the lowest (0.58). Tree biomass

Root/shoot ratio

was also analyzed by partitioning into the basic fractions of stem, crown, roots, and the
root/shoot ratio. Biomass allocation was greatest in the crown fraction and the mean root/
shoot ratio was 0.58. The carbon stock of the caldenales considering only calden tree
biomass is 20.2 Mg ha
1. While the overall carbon balance of the region is negative
(deforestation and biomass burning, the remnant forested area has increased their calden
density and in an indirect way his carbon sequestration capacity could also be increased.
2014 Elsevier Ltd. All rights reserved.

1.

Introduction

Deforestation is the major global source of carbon emission

into the atmosphere, representing 18e20% of total world

emissions [1]. Argentina, in recent years, has experienced the

16th highest absolute deforestation rate in the world and the
highest deforestation rate outside the tropics [2]. Preliminary
estimates of carbon emission from deforestation in Argentina

* Corresponding author. E.T.S. de Ingenieras Agrarias, Universidad de Valladolid, Campus de Palencia, Av. Madrid 44, C.P. 34004, Spain.
Tel.: 34 979108427; fax: 34 979108440.
E-mail address: luciarisio@gmail.com (L. Risio).
http://dx.doi.org/10.1016/j.biombioe.2014.03.038
0961-9534/ 2014 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Risio L, et al., Aboveground and belowground biomass allocation in native Prosopis caldenia
Burkart secondaries woodlands in the semi-arid Argentinean pampas, Biomass and Bioenergy (2014), http://dx.doi.org/10.1016/
j.biombioe.2014.03.038

b i o m a s s a n d b i o e n e r g y x x x ( 2 0 1 4 ) 1 e1 2

suggest this to be the countrys major source of carbon release

[3]. However, there is some evidence of forest area increase in
Argentina, specifically in northern and central forest ecosystems that were historically used for extensive grazing and are
unsuitable for modern agriculture.
Argentina is divided into seventeen ecoregions according to
climateandvegetationcharacteristics[4].Theamountofcarbon
storedinplantbiomassandthesoilorganicfractionvariesamong
ecoregions.TheEspinalecoregioncoversanareaof329,395km2and
isdividedintothreedistrictsaccordingtothedominantspecies:the
andubaydistrict,dominatedbyProsopisaffinisSpreng;theAlgarN
robodistrictdominatedbyProsopisflexuosaD.C;andthesouthernmostCaldendistrictdominatedbyProsopiscaldenia.Thisdistrictis
distributed throughout central and south of San Luis province,
southwestofCordoba,centralofLaPampa,southofBuenosAiresto
theColoradorivervalleyandnortheastofRoNegro[5e11].
With the advance of the agricultural boundary and its
ecological and environmental impact on fifteen Argentina
ecoregions over 50 years, only the Espinal has experienced a
relative increase in woody area. However, here also rapid
deforestation is causing it to function as a carbon source
rather than a carbon sink, resulting in a positive greenhouse
gas balance [12].
The Calden district houses almost all the remaining
forest area of the Espinal region [13]. Major changes
occurred in the area with the arrival of new settlers and the
railway at the end of the 19th century, and have only
intensified since then. The biggest historical threats to the
region have included the replacement of natural systems
with agriculture, overgrazing by livestock, changes in fire
regimes, extractive logging and the introduction of nonnative species [14e16]. Currently, cattle ranching are the
main economic activity in the region, followed by crop
production [17].
Although approximately 47% of the earths surface belongs
to arid and semiarid ecosystems, to date only a few studies on
biomass and carbon sequestration have been carried out in
these ecosystems [18]. This is probably due to the fact that a
significant amount of the biomass in these systems is underground, the most costly and difficult fraction to quantify
[19]. However, these ecosystems are too extensive to ignore;
changes in vegetation and carbon sequestration in these areas
must be included in discussions on global carbon balance. The
carbon sequestration potential of a forest depends on such
factors as species composition and growth rates, stand age,
disturbances (i.e herbivory, fires, harvesting), the inherent
biological capacity of the site or the use of the products obtained from the site [20].
Much of the calden woodland has been intensively
degraded and fragmented [17], loosing capacity to provide
social, economics and environmental services: they support
931 species of plants, 59 of them with potential medical
application, and 333 species of vertebrates live in the area,
many of them endemic of central Argentina [16]. For this to
change, a solution could be that carbon sequestration become
in an objective in calden management plans. International
initiatives such as REDD projects (Reducing Emission from
Deforestation and Forest Degradation in Developing Countries) offer
a financial reward, or a kind of payment for this environmental service.

There are two broad categories of processes responsible

for sources and sinks of carbon in forest regions. The first
category includes natural physiological processes (photosynthesis, respiration, decomposition, evapotranspiration),
plant and soil processes that respond to environmental
drivers, such radiation, precipitation, temperature and nutrients. The second broad category of processes falls under
the heading of ecosystem disturbances, including both direct
anthropogenic effects (e.g.: deforestation and conversion to
pasture) and natural or indirect anthropogenic effects (e.g.:
wildfire) [21].
The carbon balance in the caldenal is very complex to
establish due to the heterogeneous disturbance history
of this region which led to a fragmented patchwork of
forest patches in different conservation states that coexist
and can act like carbon sources (recently deforested areas,
burned or overmature stands) or sinks (areas invaded
by young second-growth formations, young stands, mature
stands, etc), inside a matrix of agricultural lands in constant expansion, grasslands and abandoned fields) [16].
This is a common situation in other forest ecosystems of
the world subject to the same disturbing factors e.g.: central Canadian boreal forest [22], Brazilian amazon [21], African forest [23].
To incorporate carbon sequestration as a forest management objective, it is essential to quantify as accurately as
possible the carbon stored in the different forest pools. It is
also imperative to improve on the estimations obtained from
the use of general coefficients or mean values from surrounding regions, which is the current tool for estimating
carbon content in calden woodlands. To our knowledge, there
is no equation for quantifying biomass for P. caldenia or
belowground biomass for any other Prosopis species. Viglizzo
[12] estimated the carbon stored in the Espinal soil and
biomass based on IPCC inventories [24] for similar forest biomes and Chauchard [25] developed a local and a standard
total volume equations for P. caldenia.
Given the importance of semiarid forest environments in
Argentina, the high substitution rates of these forests for
agriculture, and the opportunity to develop forest conservation strategies based on the provision of environmental services, useful silvicultural tools are needed at the operational
level. The aims of this work were to develop a biomass
quantification model for P. caldenia, determine its root/shoot
ratio and analyze the biomass partitions of the tree.

2.

Materials and methods

2.1.

Study area

Calden forests are semiarid woodlands covering about

170,000 km2 of central Argentina. This xerophytic open forest
is a transitional ecosystem between the Pampas grasslands
and the dry Monte shrublands. It is dominated by the calden
tree (P. caldenia) an endemic species in Argentina. These
woodlands thrive on the edge of the driest area of the
Argentinean Pampas, at 34e36 S and 64e66 W [26].
Calden woodlands are also composed of Prosopis nigra
Griseb, P. flexuosa (algarrobo), Celtis tala Gillies ex. Planch (Tala)

Please cite this article in press as: Risio L, et al., Aboveground and belowground biomass allocation in native Prosopis caldenia
Burkart secondaries woodlands in the semi-arid Argentinean pampas, Biomass and Bioenergy (2014), http://dx.doi.org/10.1016/
j.biombioe.2014.03.038

b i o m a s s a n d b i o e n e r g y x x x ( 2 0 1 4 ) 1 e1 2

and Geoffroea decorticans Gill. ex. H. and A. Burkart (Chanar).

The understory of perennial grasses is frequently interrupted
by dunes, wetlands and lagoons [27]. In its natural distribution
area, the climate is temperate and semiarid [28] with an
important hidric deficit in at least 6 or 7 months of the year
[29]. Total annual precipitation varies from 450 to 620 mm and
is mainly concentrated in spring and summer (SeptembereMarch). The annual isotherm range is 16e18  C.
The area consists of a well-drained plain varying from 50 to
about 600 m above sea level, with moderate slopes produced
by wind and fluvial action [13]. Soils are well drained, with low
water holding capacity and 1.5e3% organic matter [30], but
poorly developed, with scarce horizon differentiation. The
area is severely affected by wind and water erosion due to
poor soil structure [30,31].
In earlier times, the Caldenal had an open tree formation
with scattered calden trees accompanied occasionally by P.
flexuosa. Density of trees was small with no canopy closure
and there was no shrub layer [32]. Instead, one encountered
rich native grassland similar to the Spanish Dehesa.
Currently, such landscapes are rare, except in over-mature
stands with regeneration problems. The most prevalent
situation in the remaining stands is one in which a combination of fire (used systematically to stimulate herbaceous
regrowth) and inappropriate livestock management have
altered forest appearance and structure by stimulating
vegetative reproduction, increasing density, colonizing new
areas and changing allometric tree relationships. All these
factors increase heterogeneity in the Caldenal woodlands,
making it more difficult to generalize hypothesis and
results.
The study site was located on privately-owned property in
the San Luis province (Fig. 1), near the northern edge of the
natural distribution area for calden (33 3031.4700 S,
65 2332.1100 W). According to Contreras [33] in the last fifty

years this area lost around 30,000 ha of dry forest that were
convert to dryland agriculture and the water table level rose
on average at a rate of 0.15 m yr
1, with total elevation increments reaching up to 10 m in 35 years. This led to an abrupt
landscape dissection process accompanied by the appearance
of deep canyons and perennial watercourses [33] close to our
sampling site (2 km aprox.). Mean annual precipitation for the
1903e2010 period was 605 mm yr
1 (INTA-Villa Mercedes
meteorological station;
33.652
65.420 ).
On this secondary woodland, cattle ranching were the
main activity for more than seventy years and there was not
active silviculture actions on it. Fire was not used systematically as a management tool but it was present (accidentally
and on purpose). Despite its proximity to the ecotone between
calden and quebracho blanco and algarrobo negro woodlands
[6] the vegetation composition of the stand is similar to the
pure stands of calden (more than 91.7% of calden trees, data
not shown). Besides calden, P. flexuosa (algarrobo) (3.2%), Celtis
tala Gillies ex. Planch (Tala) (0.6), Geoffroea decorticans Gill. ex.
H. and A. Burkart (Chanar) (3.2%), Jodina rombifolia Hook et Arn
(0.7%) and Apidosperma quebracho-blanco Schletchtendal (0.6%).
The physiognomy of the stand can be define as a high density
forest with grassland [17].
Although, it is very difficult find two calden woodlands
alike [17] we consider that this is a representative stand of a
secondary calden woodlands which today prevail.

2.2.

Data sampling

Field work was carried out in the summer of 2011/2012

(NovembereDecembereJanuary) when trees were in active
growth and the leaves were complete developed. An inventory
was carried out on 30 plots randomly distributed over a 92 ha
area. Tree stand level variables are given in Table 1. Stand
diameter and height structures are represented in Fig. 2a and b.

Fig. 1 e Location of the sampling site and natural distribution area of Prosopis caldenia in the Argentinean Pampas.
Please cite this article in press as: Risio L, et al., Aboveground and belowground biomass allocation in native Prosopis caldenia
Burkart secondaries woodlands in the semi-arid Argentinean pampas, Biomass and Bioenergy (2014), http://dx.doi.org/10.1016/
j.biombioe.2014.03.038

b i o m a s s a n d b i o e n e r g y x x x ( 2 0 1 4 ) 1 e1 2

Table 1 e Summary of the main tree and stand level variables from Prosopis caldenia plot measurements (n [ 30).
Mean
SD
Minimum
Maximum

DBH

BD

Ht

Crown height

Crown area

Tree density. ha
1

% Single

% Multi

Age

11.51
6.51
3.52
54

16.12
8.71
7.51
57.61

4.91
1.32
1.42
11.41

1.71
0.62
0.23
5.72

4.43
5.01
0.02
54.43

453
195
180
960

48
12.92
22
69.52

52
12.91
31
78

24
4.93
14
41

Note: SD, standard deviation; DBH, diameter at breast height (cm); BD, diameter at tree base (cm); Ht, tree height (m); Crown height, crown base
height (m), Crown area in m; % single, percentage of single-stemmed trees; % multi., percentage of multi-stemmed trees (usually 2e3 stems per
tree); Age in years.

Plot dimensions were identical to those used in the First

National Inventory of Argentinean Native Forest (NFI1) [13]. In
each plot, every calden tree with a basal diameter greater than
7.5 cm was recorded and the following dendrometric variables
were measured: diameter at breast height (DBH 1.3 m above
ground) in cm, diameter at tree base (DB) in cm, tree height in
m, crown base height in m, point of maximum crown width in
m, bifurcation height in cm and maximum and minimum
crown width in cm. Diameters were measured with calipers,
heights were measured with a VERTEX III digital hypsometer
and crown width was measured with a tape measure.
Eleven diameter classes were established in 5 cm increments; with the first four (smallest) classes accounting for
almost 92% of the stand population and the 73% of total basal
area. The minimum recorded diameter at the base was 7.5 cm
and the maximum 57.61 cm. Of a total of 629 inventoried trees,
580 showed a BD between 7.5 and 27.49 cm. Consequently, the
38 sample trees were randomly selected from among these
four classes (Table 2).
The selected trees were felled and divided into the
following sections: leaves, roots, stem with bark, branches
less than 2 cm in diameter, branches 2e7 cm in diameter and
branches greater than 7 cm in diameter (Table 2). The complete root systems were extracted with a bulldozer (FIAT AD
18); except for the thinnest root fractions, which could not be
captured and were therefore not included in the analysis. The
depth root extraction varies between 1 and 3.5 ms according to
the tree size.
All biomass fractions were weighed in the field with a
portable Wei Heng model A05L scale (10 g precision) to obtain
fresh weight by fractions. Representative samples of each
fraction were taken to the laboratory, where they were oven
dried at 80  C until reaching a constant weight. From this we
obtained estimates of moisture content and dry weight. A
cross section of 5 cm of width at the base of the stem were cut
for dendrochronological studies.
Cross-sections were polished and then dated under a microscope following the methodology described by Stokes and
Smiley [34]. Two radii were dated per cross-section. Tree rings
were measured to 0.01-mm accuracy with a measurement
system connected to a computer. Cross-dating and measuring
quality were checked with the program COFECHA [35].
The biomass distribution within the tree and the root/
shoot ratio were also calculated and analyzed.

2.3.

Data analysis

Data analysis consisted of four consecutive steps: a) correlation analysis b) individual fitting of the evaluated biomass

equations; c) simultaneous fitting of all the biomass equations

adjusted for each fraction; and d) partition and distribution
analysis of the tree biomass.
First of all, a correlation analysis was carried out between
all the tree variables and each biomass fraction in order to
determine which independent variables were more correlated
with each biomass component.
Secondly, different lineal and non-lineal equations (Table
3) were evaluated to relate dry biomass with different tree
variables. Each equation was individually fitted and analyzed,
based on the graphical analysis of residuals. The following
statistics parameters were used to identify the most appropriate equation: root of the mean quadratic error (RMSE), sum

Fig. 2 e a). Number of stems for each diameter classes of

the Prosopis caldenia plots (n [ 30). (Class 1: 7.5e11.49 cm;
class 2: 11.5e16.49 cm; class 3: 16.5e21.49 cm; class 4:
21.5e26.49 cm; class 5: 26.5e31.49 cm; class 6:
31.5e36.49 cm; class 7: 36.5e41.49 cm; class 8:
41.5e46.49 cm; class 9: 46.5e51.49 cm; class 10:
51.5e56.49 cm; class 11: 56.5e61.49 cm). b). Number of
stems for each height classes of the Prosopis caldenia plots
(n [ 30). (Class 1: 1.4e1.99 m; class 2: 2e2.99 m; class 3:
3e3.99 m; class 4: 4e4.99 m; class 5: 5e5.99 m; class 6:
6e6.99 m; class 7: 7e7.99 m; class 8: 8e8.99 m; class 9:
9e9.99 m; class 10: 10e10.99 m; class 11: 11e11.99 m).

Please cite this article in press as: Risio L, et al., Aboveground and belowground biomass allocation in native Prosopis caldenia
Burkart secondaries woodlands in the semi-arid Argentinean pampas, Biomass and Bioenergy (2014), http://dx.doi.org/10.1016/
j.biombioe.2014.03.038

b i o m a s s a n d b i o e n e r g y x x x ( 2 0 1 4 ) 1 e1 2

Table 2 e Standard deviation, maximum, minimum and mean values of tree level variables and the different tree biomass
fractions (n [ 38) for Prosopis caldenia.
Total height

DBH

BD

AB

Dry biomass fractions

Branches

Mean
SD
Minimum
Maximum

5.21
1.02
3.01
7.51

11.03
3.92
5.01
21.12

14.31
4.72
7.53
24.92

181.38
114.27
44.53
487.83

<2

2e7

>7

7.22
7.31
1.71
39.93

8.01
5.51
0.22
22.01

10.93
10.11
0.00
42.7

Leaves

Stem

Roots

Total

1.41
1.02
0.34
4.40

11.82
10.44
1.61
49.32

20.31
15.42
3.31
66.42

59.52
40.83
10.12
199.91

Note: SD, standard deviation; Total height in m; DBH, diameter at breast height (cm); BD, diameter at tree base (cm); AB, area of the base of the
stem; <2, branches less than 2 cm in diameter (kg); 2e7, branches 2e7 cm in diameter (kg); >7, branches greater than 7 cm in diameter (kg);
Total, total biomass (kg).

Table 3 e Biomass models evaluated for different tree components for Prosopis caldenia.
Model
1
2
3
4
5
6

Equation
W
W
W
W
W
W

Model

b*(AB*Ht)
b*(AB2*Ht)
(b*AB) (l*AB2) (q*AB2*Ht)
(b*AB) (l*Ht)
(b*AB2) l*(AB2*Ht)
a b*(AB2*Ht)l

Equation

7
8
9
10
11

W
W
W
W
W

(b*AB2) (l*Ht)
(b*AB2) (l*Ht) (q*AB2*Ht)
(b*AB2) l*(AB*Ht)
b*(AB2*Ht) l*(AB*Ht)
b*(ABl)*(Htq)

Note. W, dry Biomass (kg); AB, area of the base (cm); Ht, tree height (m); a, b, q, l, parameters of the models.

3.

Results

The dendrochronological analysis showed that the maximum

age of the selected trees was 41 and the minimum 14 with a
mean of 24 years (Table 1). The age structure of the analyzed
trees is showed in Fig. 3.
The area of the base of the stem (AB) showed the highest
correlation value with dry biomass in all the fractions. BD and
DBH also performed well. Only tree AB and BD were statistically significant. Crown height showed the lowest correlation
values, while tree height and crown diameter ranked fourth or
fifth, depending on the fraction considered (Table 4).
Model 7 showed the best independent fit statistics and all
estimated parameters were significant for all three biomass
fractions. All evaluated model presented a poor performance
with the leaves fraction, so based on these results, this

18
16
14

Number of trees

12
10
8
6
4
2
14

21

23

24

25

26

31

39

0
41

of square of error (SSE), adjusted correlation coefficient R2_adj.

and significance of the estimated parameter (p-value).
Eleven models that used the area of the base and tree
height as independent variables were evaluated for each
biomass fraction.
The next step consisted of producing a simultaneous fitting
of the equation system. The best models of each fraction were
simultaneously fitted using joint generalized regression
(Seemingly Unrelated Regressions, SUR) [36] to ensure the additivity property between biomass components [37,38]. This
method includes cross-equation error correlations, in order to
obtain the sum of total aerial biomass components. Weighted
regression was used to avoid heteroscedasticity: each observation was weighted by the inverse of its variance to homogenize the variance of residuals. This weighting factor was
estimated using a power function of an independent variable
[37,39]. Multicollinearity was tested using the condition
number. The MODEL procedure in the SAS/ETS software was
used to obtain model fits [40].
The root/shoot ratio between total aboveground and
underground biomass was calculated and biomass distribution within the tree was also analyzed. To simplify the
analysis, trees were divided into three components: crown
(formed by leaves and branches of all diameters), stem and
root system.
With the fitted models and the inventory data, the partitioning of tree biomass into crown, stem and roots, total and
belowground biomass and total and belowground carbon
stock per ha were estimated and analyzed. The biomass
values were converted to carbon stocks with a coefficient
value equal to 0.48 which is the average value of the P. affinis
and P. nigra coefficients [41] since there is not a specie specific
value.

Years

Fig. 3 e Age structure of the Prosopis caldenia fractioned

trees. (n [ 38).

Please cite this article in press as: Risio L, et al., Aboveground and belowground biomass allocation in native Prosopis caldenia
Burkart secondaries woodlands in the semi-arid Argentinean pampas, Biomass and Bioenergy (2014), http://dx.doi.org/10.1016/
j.biombioe.2014.03.038

b i o m a s s a n d b i o e n e r g y x x x ( 2 0 1 4 ) 1 e1 2

Table 4 e Correlation coefficients between biomass component and dendrometric variables of Prosopis caldenia.
Branches <2

Correlation coeff.

Branches 2e7

Correlation coeff.

Branches>7

Correlation coeff.

Leaves

Correlation coeff.

Roots

Correlation coeff.

Stem

Correlation coeff.

Total

Correlation coeff.

Hc

Ht

DBH

DB

AB

Dc

0.09
(0.6213)
0.01
(0.9711)

0.02
(0.9123)
0.03
(0.8834)
0.01
(0.9612)
0.07
(0.7022)

0.02
(0.9032)

0.11
(0.5231)
0.34
(0.0443)
0.24
(0.1542)
0.14
(0.4014)
0.29
(0.0833)
0.24
(0.1522)
0.33
(0.0412)

0.48
0.003
0.63
<0.0001
0.67
<0.0001
0.56
0.0004
0.66
<0.0001
0.61
<0.0001
0.56
0.0004

0.64
(<0.0001)
0.77
(<0.0001)
0.76
(<0.0001)
0.72
(<0.0001)
0.87
(<0.0001)
0.82
(<0.0001)
0.83
(<0.0001)

0.67
(<0.0001)
0.77
(<0.0001)
0.77
(<0.0001)
0.74
(<0.0001)
0.89
(<0.0001)
0.85
(<0.0001)
0.86
(<0.0001)

0.17
(0.3333)
0.06
(0.7523)
0.14
(0.4111)
0.19
(0.2514)
0.13
(0.4400)
0.11
(0.5323)
0.09
(0.6101)

Note: p-value between parentheses, Hc, crown height; Ht, tree total height; DBH, diameter at breast height; DB, tree basal diameter; AB, area of
the base; Dc, tree crown diameter.

fraction was combined with that of branches less than 2 cm in

diameter, resulting in a significant improvement in model
behavior (Table 5).
The model that provided the best fit for each independent
component was later included in the SUR fitting. All calculated
model parameters were statistically significant at the 99%
confidence level. Stem and total biomass showed the highest
R2_Adj. value and branches over 7 cm in diameter the lowest
(Table 6).
With the inventory data and the adjusted models total and
aboveground biomass in each plot were calculated. For total
biomass the mean value was 20.2, the maximum 69.8 and the
minimum 4.1 Mg ha
1 of dry matter. Mean aboveground
biomass was 11.5, maximum 31.5 and minimum 2.6 Mg ha
1
of dry matter. The mean total carbon stock per ha was 9.6, the
maximum 33.3 and the minimum 1.9 Mg ha
1. In the case of
aboveground carbon stock the mean value was 5.5, the
maximum 15.0 and the minimum of 1.3.

The partitioning of tree biomass into basic fractions such

as the stem with bark, the crown (foliage all branches) and
the root system is shown in Fig. 4. The P. caldenia tree allocates
more biomass to the crown than to the stem or root system,
with a mean value of 46% of total tree biomass, a minimum of
22% and a maximum of 82%.
The root system receives the second largest biomass allocation, with a mean value of 35%, a maximum of 65% and a
minimum of 6%. The smallest biomass allocation for P. caldenia corresponds to the stem fraction, with a mean value of
19%, a maximum of 38% and a minimum of 4% of total tree dry
biomass. The P. caldenia root/shoot ratio presented a mean
value of 0.58, with a maximum of 1.88 and a minimum of 0.07.
The relations between the root/shoot ratio and the three
age (Fig. 5) and between root/shoot ratio and three stand
density (Fig. 6) were also analyzed.
The inventory data and the fitted models were used to
estimated the partitioning of tree biomass along an AB

Table 5 e Independent adjustments of the evaluated models with significant estimated parameters for each biomass
fraction of Prosopis caldenia.
Fraction
Branches <2 cm leaves
Branches 2e7 cm
Branches >7 cm
Stem

Roots

Total biomass

Model

SSE

RMSE

R2_Adj.

Model

RMSE

R2_Adj.

1
2
1
2
1
2
1
2
6
1
2
6
1
2
6

1410.12
1668.41
764.01
1178.92
1962.12
2509.72
2421.91
2580.72
1600.52
5706.91
7137.12
3533.43
41,001.61
54,132.21
22,836.31

6.17
6.72
4.54
5.65
7.28
8.24
8.09
8.35
6.67
12.42
13.89
9.36
33.29
38.25
25.19

0.05

0.13
0.30

0.08
0.34
0.16
0.40
0.36
0.59
0.36
0.20
0.59
0.38
0.19
0.65

302.11

3.03

0.71

6
7
6
7
7
8
9
7
9

480.14
304.74
1421.93
476.91
625.61
971.32
2264.21
2341.52
5594.01

3.65
3.09
6.29
4.06
4.42
5.27
7.93
8.84
12.47

0.55
0.66
0.51
0.76
0.82
0.75
0.43
0.70
0.36

7
9

13,465.53
40,412.21

21.19
33.51

0.78
0.38

SSE

Note: Branches <2, branches less than 2 cm in diameter; Branches 2e7, branches 2e7 cm in diameter; Branches >7, branches greater than 7 cm
in diameter; SSE, sum of square error; RMSE, root of the mean quadratic error; R2_Adj., r2 adjusted correlation coefficient. All estimated parameters were significant at 0.0001.

Please cite this article in press as: Risio L, et al., Aboveground and belowground biomass allocation in native Prosopis caldenia
Burkart secondaries woodlands in the semi-arid Argentinean pampas, Biomass and Bioenergy (2014), http://dx.doi.org/10.1016/
j.biombioe.2014.03.038

b i o m a s s a n d b i o e n e r g y x x x ( 2 0 1 4 ) 1 e1 2

Table 6 e Simultaneous fit of biomass equations for Prosopis caldenia.

R2_adj.

Power values

Selected model

Estimated parameters

9.14

0.71

1.22988*10
7

W (b*AB2) (l*h)

333.82

9.82

0.64

1.71880

W (b*AB2) (l*h)

1049.72

30.87

0.58

2.39614

W (b*AB2) (l*h)

Stem

957.14

28.15

0.75

2.16213

W (b*AB2) (l*h)

Roots

2891.51

85.04

0.65

6.78854

W ( b* AB2) (l*h)

16,982.71

566.12

0.75

1.66889

W (b*AB2) (l*h)

0.000054
0.784828
0.000045
0.948952
0.00007
1.108615
0.00009
1.123
0.000142
2.171142
0.000401
6.136537

Biomass fraction

SSE

Branches <2 leaves

310.83

Branches 2e7
Branches > 7

Total biomass

RMSE

Note: SSE, sum of error square; RMSE, root of the mean quadratic error; R2_Adj., r2 adjusted correlation coefficient; Power values, factor used in
the power function to correct heteroscedasticity.

gradient (Fig 7). Only the data between the diameter range
used to fit the equation system (DBH between 7.5 and 24.9)
were used. The AB varies from 0.2 to 3.4 m2 ha
1, the crown
fraction had a mean percentage of 45.40, a maximum of 46.28
and a minimum of 42.82, the stem mean value was 19.21, the
maximum 21.77 and the minimum 18.34 and finally the roots
fraction showed a mean value of 35.39, a maximum of 35.41
and a minimum of 35.38.

4.

Discussion

Estimation of forest carbon stocks from forest inventories

requires accurate biomass models that are simple enough to
provide policy-makers and stakeholders with comparable and
verifiable information for different areas. This article

describes the results of a study to test a method developed for

quantifying above and below-ground biomass allocation in
native P. caldenia woodlands of the semi-arid Argentinean
pampas. The applied methodology allowed us to quantify
total dry biomass with available data in an operational
process.
Allometric equations for P. flexuosa showed differing
annual radial increments between single and multi-stemmed
individuals, possibly due to differences in the photosynthetic
capacity of growth patterns [42]. Beyond a certain stem
diameter, crown area decreased in multi-stemmed trees due
to competition between stems of the same tree, reversing the
initial growth trend. According to our inventory data (Table 1),
for calden almost the half of the trees is single and the other
multi-stemmed trees. These percentages could vary according
to the stand management history.

100
90

Pe rce nt aje of t ot al dry biomass by fract ion

80
70
60
Crown

50

Stem
40

Roots

30
20
10

0.43
0.44
0.50
0.50
0.60
0.66
0.81
0.83
0.83
0.90
0.96
1.01
1.02
1.03
1.18
1.26
1.28
1.28
1.30
1.36
1.41
1.50
1.67
1.74
1.76
1.97
1.99
2.00
2.32
2.34
2.52
2.88
3.45
3.46
3.60
4.03
4.62
4.89

Area of the stem base (m 2 )

Fig. 4 e Real total dry biomass partitioning of Prosopis caldenia felled trees (n [ 38): crown (all branches plus foliage), stem
and roots.
Please cite this article in press as: Risio L, et al., Aboveground and belowground biomass allocation in native Prosopis caldenia
Burkart secondaries woodlands in the semi-arid Argentinean pampas, Biomass and Bioenergy (2014), http://dx.doi.org/10.1016/
j.biombioe.2014.03.038

b i o m a s s a n d b i o e n e r g y x x x ( 2 0 1 4 ) 1 e1 2

Fig. 5 e Variation of the root/shoot mass ratio versus tree

age of Prosopis caldenia trees (n [ 24).

Fig. 6 e Variation of the Root/Shoot mass ratio versus tree

stand density of Prosopis caldenia (n [ 24).

In the stand where samples were taken the three main

disturbing factors that modify the growth patterns identified
for the caldenal acted: herbivory, logging and fire. Cows and
sheep grazed on this stand, these two animals have different

effects on tree population, while sheep farming was not

associated with increases in calden densities (sheep consumes calden seedlings), high recruitment of calden occurred
following the introduction of cattle (cows eat the fruit and
break the seed dormance) [43]. Firewood extraction for subsistence was also practice in this stand, quantify the amounts
or percentage of the basal area extracted is impossible to
estimate.
Finally, anthropogenic and wild fires were present; forest
fires are a recurrent and important part of the disturbing
factors that affect the calden woodlands [44e47]. Several authors pointed it as the leading factor to the high tree density
increment in the caldenales [15,48e50] since, calden have the
capacity of resprout after been affected for a fire [15] and
favored the implantation by triggering dormant seed germination [43]. The effect of fire on tree growth varies according to
the percentage of cambial perimeter damaged, Ryan [51] affirms that a damage over than 20e25% of the cambial perimeter could seriously compromises the tree survival
possibilities to a forest fire. In some Prosopis species, damages
over the 50% of the cambial perimeter in the main branches
were observed until 7 ms high, fire also produces structural
variations in the stem cells disorganizing the post fire axial
tissues, reducing tree vitality and favoring pathogens attacks
[52].
Two fires were necessary to alter the vegetation structure
and modify the cover-biomass ratio from a linear relationship
in control areas (no fire) to a nonlinear relationship in burned
areas [53]. Morphological characteristics of plant regeneration
after a fire include lower total height and a greater below/
aboveground biomass ratio than unburned trees [54].
Since almost all the Caldenal had suffered different levels
of disturbance [17] that modifies the allometric relationships
and the number of stems, we used has input variables the area
of the base of the stem and tree high. The area of the base was
calculated from the basal diameter of the stump, regardless of
the number of aerial stems. Tree height was included as an
explanatory variable in all models, though in some fractions
its inclusion did not significantly improve the fitted model.

Fig. 7 e Estimated total dry biomass partitioning of Prosopis caldenia trees (n [ 520): crown (all branches plus foliage), stem
and root system.
Please cite this article in press as: Risio L, et al., Aboveground and belowground biomass allocation in native Prosopis caldenia
Burkart secondaries woodlands in the semi-arid Argentinean pampas, Biomass and Bioenergy (2014), http://dx.doi.org/10.1016/
j.biombioe.2014.03.038

b i o m a s s a n d b i o e n e r g y x x x ( 2 0 1 4 ) 1 e1 2

Tree height was included because it is present in forest inventories and indirectly facilitates information about the
competitive environment in which the tree was developed:
such as stand age, density or site quality [55]. This makes the
model less specific and more adaptable for use at different
sites [56].
In foliar biomass estimates for P. flexuosa based on allometric relations [57], estimations using tree level variables
measured in the crown and at the base of the live crown
showed a better fit than those based on stem measurements.
Foliar biomass estimations for P. flexuosa using basal diameter
as a predictor gave the worst results, followed by estimations
based on measurements at the stem apex.
Similar behavior was observed in P. caldenia: foliar biomass
predictions based on tree basal area, basal diameter or
diameter at breast height were inadequate and were not
improved by incorporating other variables such as tree height.
To solve this problem, the foliar fraction was combined with
the fraction for branches less than 2 cm in diameter. These
two fractions are closely linked and would have the same use
in a harvesting plan; therefore their homogenization presents
no major practical problems. All models of the equation system presented a similar independent setting, with a minimum
R2_adj. of 0.66 for branches 2e7 cm in diameter and a
maximum of 0.82 for stem with bark (Table 5).
The partitioning of biomass within the tree indicates that
calden allocates the greatest part of its resources to crown
development (leaves plus branches of all diameters), then to
the root system and finally to the stem (Fig. 4). Similar biomass
partitioning proportions have been described for Quercus ilex L.
and Quercus suber L. and other hardwood species in the Iberian
peninsula [58]. P. caldenia tends to ramify and bifurcate to low
high [17], these stem bifurcations may have been considered
has branches of more than 7 cm of diameter, which may have
led to this biomass partitions values. Besides, calden is an
heliofita species [17] so it is natural that allocate a big part of
its aboveground biomass in the crown.
The biomass partitioning within the tree obtained from the
inventory data and the estimated models do not reflect the
real variability of the biomass partitioning within the tree
(Fig. 7). The obtained values are almost equal to the real mean
values: 19% for the stem, 35% for the root system and 46% for
the crown. This issue has to be considered since these three
basic components of the tree led to different uses and products and they proportions can vary according to the management history of the stand (fire, herbivory, wood extraction,
pruning).
According to the IPCC biome classification [24], the calden
woodlands are temperate continental forest, and the total
aboveground biomass estimated (included woody and herbaceous species) for these woodlands are equal to 60 Mg ha
1
(10e130) for 20 years old forest or 130 Mg ha
1 (50e200) for
forest older than 20 years, the values between parentheses are
the intervals. We estimated the dry matter per hectare for a
calden woodland with an average age of 24 years (we not
consider other tree, brushes or herbaceous species) and our
mean value estimated is equal to 11.5 Mg ha
1 (2.6e31.5).
Our value is lower than that of IPCC, but part of this variability could be due for the unanalyzed compartments. However this probes the importance of determinate specific

coefficient for each ecosystem to reduce the global uncertainties in the global carbon balance. Viglizzo and Jobaggy
[12] analyzed the balance and the carbon stock in the Espinal
for the period of 1956e2005 and they found a slight increase of
the woody area which agree with data from Dussart [43]
mainly due to the fire and cattle breeding. However, they
concluded that this region has been acting like a source and
not a like a pool of carbon due to the high deforestation rate
and the subsequent biomass burning.
The calden root/shoot ratio varies from 1.88 to 0.07.
Mokany [59] observed that for forest and woodlands, root/
shoot ratios were related to several factors associated with
stand development. They decreased significantly with increases in shoot biomass, stand age, stand height, and mean
DBH, and increased with tree density. For P. caldenia we do not
observe a clear trend for root/shoot ratios as a function of tree
age (Fig. 5), but there is one between the root/shoot ratio and
tree density (Fig. 6). This could be associated to the fire history
of the stand, since fire may stimulate calden implantation
rates and change the structural pattern of growth when only
kills the aboveground part of the tree, stimulating the
regrowth of stem bases [43].
Other abiotics factors that could affect the root/shoot ratio
are: soil characteristics, as nutrient availability [60,61] and
texture [62,63] or the water table level, since decreasing soil
moisture produces higher root/shoot ratios [64] and P. caldenia
can have a groundwater consumption with shallow water
table levels [65]. Although soil types, water table level and fire
scarce in tree rings were not checked in this work.
Channel [66] and Cairns [67] found no significant differences in the below/aboveground biomass ratio between
different groups of species and established a general ratio of
0.25 for deciduous species, regardless of latitude or soil
texture. However, we found a root/shoot mean ratio for P.
caldenia equal to 0.58. This value coincides with the results of
Ruiz-Peinado for Spanish hardwood species [58], who calculated the root/shoot ratios for the main species of the Iberian
Peninsula and found that their mean values did not coincide
with that of Cairns [67] either. Several species showed similar
or even higher mean ratios (in parentheses) than that of calden: such as Quercus canariensis Willd (0.49), Quercus faginea
Lamb (0.36), Castanea sativa Miller (0.77), Ceratonia siliqua L.
(0.81) or Alnus glutinosa (L.) Gaertn (0.81).
The differences between Iberian or calden tree values and
that established by Cairns [67] can partly be explained by
management practices (harvesting of aerial parts for firewood
or charcoal extraction and burning to stimulate herbaceous
regrowth), but are mainly due to the arid/semi-arid climate, in
which deep and extensive root systems are better suited for
water and nutrient uptake. Gaston [68] determined an average
ratio of 0.47 for forests and savannas in arid and semi-arid
climate, which is closer to our results.
The natural heterogeneity of the caldenal is heightened for
some disturbance factors and this is a disadvantage for the
model behavior. However, a great part of this created variability is indirectly collected by the selected input variables:
fire and cattle breeding are related with an increased in calden
tree disperse, recruitment, survive and resprout which
determinate the stand density [60] and this one, modifies the
tree growth in height and the root/shoot ratio. Fire also

Please cite this article in press as: Risio L, et al., Aboveground and belowground biomass allocation in native Prosopis caldenia
Burkart secondaries woodlands in the semi-arid Argentinean pampas, Biomass and Bioenergy (2014), http://dx.doi.org/10.1016/
j.biombioe.2014.03.038

10

b i o m a s s a n d b i o e n e r g y x x x ( 2 0 1 4 ) 1 e1 2

modifies tree growth in height and the resprouting capacity

stimulating the regrowth from stem base leading to multistemmed trees. Thought the use of the area of the base the
number of stems is not necessary to consider. In the volume
equations developed for calden [25] a close relationship between volume and basal area regardless in the number of
aerial stems was found.
Finally, it is important to note that accurate estimation of
woody biomass requires knowledge of the stand evolution,
the land tenure regime and the history of disturbances [69].
These can be critical factors, since they modify allometric tree
relationships. Reducing data variability could significantly
improve the equation systems developed for P. caldenia. An
example of this would be to incorporate stratified sampling
based on stand disturbance into the models as an explanatory
variable.

[2]
[3]

[4]

[5]
[6]

[7]

5.

Conclusions
[8]

The equation systems presented here made it possible to more

precisely and accurately estimate the biomass stored in calden
woodlands. Wood production is not the main function in these
forests, so biomass estimations by fractions are highly useful for
identifying the best forest management practices and understanding the role of the forest as a carbon sink.
The selected input variables indirectly collected the caldenal heterogeneity and the effects of different disturbance
factors that modify the tree allometric relationships, this
make the model less specific and more adaptable for use at
different sites.
The models fitted could be applied to NFI1 data, which
records the diameter and height of all sample trees per plot.
This would facilitate carbon counts at the ecosystem, regional
and national levels.

[9]

[10]
[11]
[12]

[13]

[14]

Acknowledgments
Comments raised by two anonymous referees have greatly
improved the presentation and accuracy of this paper. We
would like to thank all those who participated in the laborious
field work, especially L. Garay, M. Nievas, P. Rossi, E. Orqun,
M. Zavala, O. Allione and C. Bintana. Thanks also to M.
Marchissio and to M. Chaparro for allowing us to participate in
his work. To Catalina Roa for the dendrochronological analysis. Finally, we gratefully acknowledge funding from the
ERASMUS MUNDUS ECW 2009 1655/001-001 European Union
mobility programme (ERASMUS MUNDUS 2009 1655/001-001
ECW) fellowship awarded to the corresponding author and
PICT2011 0745 MinCyT for economical support in the sampling
activities.

references

[15]

[16]

[17]

[18]

[19]
[20]

[1] Solomon S, Qin D, Manning M, Chen Z, Marquis M,

Averyt KB, et al. Contribution of Working Group I to the
Fourth Assessment Report of the Intergovernmental Panel

on Climate Change 2007. Cambridge (UK): Cambridge

University Press; 2007.
FAO. State of worlds forest 2007. Rome (Italy): Food and
Agriculture Organization of the United Nations; 2007. p. 141.
Gasparri NI, Manghi E. Estimacion de volumen, biomasa y
contenido de carbono en las regiones forestales Argentinas.
Informe final. Buenos Aires: Direccion de bosques. Secretara
de ambiente y desarrollo sustentable; 2004.
Sistema de Informacion Nacional Ambiental. Mapas
tematicos nacionales, Ecorregiones [Map on the internet].
Buenos Aires (Argentina): Secretara de ambiente y
desarrollo sustentable de la nacion Argentina. [cited 2013
August 15]. Available from: http://www.ambiente.gov.ar/?
aplicacionmapoteca&idseccion76&IdApli2
Lasalle JC. Informaciones descriptivas de los Caldenales.
Rev For Argent 1966;10(1):15e9.
guila J, Bernardon AE. Las formaciones
Anderson D, Del A
vegetales en la provincia de San Luis. Rev Investig Agropecu
INTA 1970;7(3):153e8.
Cabrera A. Regiones Fitogeograficas Argentinas. Enciclopedia
Argentina de Agricultura y Jardinera Tome II, Fascicle 1. 2nd
ed. Buenos Aires (Argentina): ACME; 1976.
Cabrera A, Willink A. Biogeografa de America Latina. 2nd ed.
Washington D.C (EEUU): OEA; 1980.
Cano E. Inventario Integrado de los Recursos Naturales de la
Provincia de La Pampa. Buenos aires (Argentina): INTAMinisterio de Economa y Asuntos Agrarios-Univ. Nac. de La
Pampa; 1980.
Lewis J, Collantes M. El espinal periestepico. Ciencia e Invest
1973;29:360e77.
Morello J. El punto de vista ecologico y la expansion pecuaria.
Ciencia e Invest 1975;31(7e9):168e80.
Viglizzo E, Jobbagy E. Expansion de la Frontera Agropecuaria
en Argentina y su Impacto ecologico-ambiental en
Argentina. Buenos Aires (Argentina): I.N.T.A; 2010.
Menendez J, La Rocca S. Primer Inventario de Bosques
Nativos. Inventario de Campo de la Region del Espinal.
andubay. Manual de Teledeteccion.
Distritos Calden y N
Buenos Aires (Argentina): Secretara de ambiente y
desarrollo sustentable de la nacion Argentina; 2007.
Amieva E, El Parque Luro, Santa Rosa: Fondo Editorial
Pampeano, APN. Las aereas protegidas de la Argentina,
herramienta superior para la conservacion de nuestro
patrimonio natural y cultural. Buenos Aires: Administracion
de Parque Nacionales; 2007.
Medina A. Reconstruccion de los regmenes de fuego en un
bosque de Prosopis caldenia, provincia de La Pampa,
Argentina. Bosque 2007;28(3):234e40.
Menendez J, La Rocca S. Primer inventario nacional de
bosques nativos. Segunda etapa, inventario de campo de la
andubay. Buenos
region del Espinal: Distritos del Calden y N
Aires (Argentina): Secretara de ambiente y desarrollo
sustentable de la nacion Argentina; 2007.
Menendez J, La Rocca S. Primer inventario nacional de
bosques nativos. Segunda etapa, inventario de campo de la
andubay. Anexo I:
region del espinal: Distritos del Calden y N
Estado de conservacion del Caldenal. Buenos Aires
(Argentina): Secretara de ambiente y desarrollo sustentable
de la nacion Argentina; 2007.
Bonino E. Changes in carbon pools associated with a landuse gradient in the Dry Chaco, Argentina. For Ecol Manag
2006;223(1e3):183e9.
Lal R. Carbon sequestration in dryland ecosystems. Environ
Manag 2004;33(4):528e44.
Pardos J. Los ecosistemas forestales y el secuestro de carbono
ante el calentamiento global. Madrid (Espana): Instituto
Nacional de Investigacion y Tecnologa Agraria and
Alimentaria Ministerio de Ciencia e Innovacion; 2010.

Please cite this article in press as: Risio L, et al., Aboveground and belowground biomass allocation in native Prosopis caldenia
Burkart secondaries woodlands in the semi-arid Argentinean pampas, Biomass and Bioenergy (2014), http://dx.doi.org/10.1016/
j.biombioe.2014.03.038

b i o m a s s a n d b i o e n e r g y x x x ( 2 0 1 4 ) 1 e1 2

[21] Potter C, Klooster S, Genovese V. Carbon emissions from

deforestation in the Brazilian Amazon Region.
Biogeosciences 2009;6(11):2369e81.
[22] Bond-Lamberty B, Peckham SD, Ahl D, Gower ST. Fire as the
dominant driver of central Canadian boreal forest carbon
balance. Nature 2007;450(7166):89e93.
[23] Ciais P, Bombelli A, Williams M, Piao S, Chave J, Ryan C, et al.
The carbon balance of Africa: sntesis of recent research
Studies. 2011. Phil Trans Math Phys Eng Sci
1943;369:2038e57.
[24] Eggleston H, Buendia L, Miwa K, Ngara N, Tanabe K. Revised
2006 IPCC Guidelines for National Greenhouse Gas
Inventories Programme. Hayama (Japan): IGES; 2006.
[25] Chauchard C, Sbrancia L, Medina A, Rabino A. Prosopis
caldenia [Burk] total volume functions, Argentina.
Quebracho 2009;17(1,2):41e51.
guila JA, Bernardon AE. Las formaciones
[26] Anderson DL, Del A
vegetales en la provincia de San Luis. Revista Investig
Agropecu INTA 1970;7(3):153e8.
[27] Cabrera AL. Regiones fitogeograficas Argentinas. Tome II,
Fascicle 1. 2nd ed. Buenos Aires (Argentina): ACME; 1994.
[28] De Fina A, Ravelo A. Climatologa y Fenologa agrcolas.
Buenos Aires (Argentina): EUDEBA; 1975.
[29] Barbosa O. Descripcion del ecosistema. In: Barbosa O,
Privitello L, editors. Caracterizacion Ecologica y utilizacion
del caldenal de San Luis (Argentina). San Luis (Argentina):
FICES; 2005.
[30] Pena Zubiate CA, Anderson D, Demmi MA, Saenz JL,
DHiriart A. Carta de suelos y vegetacion de la provincia de
San Luis. Buenos Aires (Argentina): Secretara de Agricultura,
Ganadera, Pesca y Alimentacion, INTA y Gobierno de la
provincia de San Luis; 1998.
[31] Cano E, Fernandez B, Vegetacion Montes DM. Inventario
integrado de los recursos naturales de la provincia de La
Pampa. Buenos Aires (Argentina): INTA, Provincia de la Pampa
and Universidad Nacional de La Pampa. Buenos Aires; 1980.
[32] Koutche J, Carmelich V. Contribucion al conocimiento de los
bosques de la Republica Argentina: Estudio Forestal del
Calden. Bol Min Agric Nac 1936;37(1e4):1e22.
[33] Contreras S, Santoni C, Jobbagy E. Abrupt watercourse
formation in a semiarid sedimentary landscape of central
Argentina: the roles of forest clearing, rainfall variability and
seismic activity. Ecohydrology 2013;6(5):794e805.
[34] Stokes M, Smiley T. An introduction to tree-ring dating.
Chicago, IL: University Chicago Press; 1968.
[35] Holmes RL. Computer-assisted quality control in tree-ring
dating and measurement. Tree Ring Bull 1983;43:69e75.
[36] Zellner A. An efficient method of estimating seemingly
unrelated regressions and tests for aggregation bias. J Am
Stat Assoc 1962;57(298):348e68.
[37] Parresol BR. Additivity of nonlinear biomass equations. Can J
For Res 2001;31(5):865e78.
[38] Bi HQ, Turner J, Lambert MJ. Additive biomass equations for
native eucalypt forest trees of temperate Australia. Trees
2004;18(4):467e79.
lvarez[39] Balboa-Murias MA, Rodrguez-Soalleiro R, Merino A, A
Gonzales JG. Temporal variations and distribution of carbon
stocks in aboveground biomass of Radiata pine and maritime
pine pure stands under different silvicultural alternatives.
For Ecol Manag 2006;237(1e3):29e38.
[40] SAS Institute INC. SAS 9.1 ETL studio: users guide. Cary, NC:
SAS Institute Inc; 2004.
[41] De Petre A, Karlin U, Ali S & Reynero N. Proyectos de
Investigacion Aplicada a los Recursos Forestales Nativos.
Alternativas de sustentabilidad del bosque nativo del
rea Captura de Carbono. [Project final report on the
Espinal. A
Internet] Buenos Aires (Argentina): Secretara de ambiente y
desarrollo sustentable de la nacion Argentina. [updated 2013

[42]

[43]

[44]

[45]

[46]

[47]
[48]
[49]

[50]
[51]

[52]

[53]

[54]

[55]

[56]

[57]

[58]

[59]
[60]

[61]

[62]

11

July 29; cited 2013 February 14] Available from: http://www.

ambiente.gov.ar/archivos/web/PBVyAP/File/A3/PIARFON%
20MyE/Captura%20de%20carbono.pdf
lvarez JA, Villagra PE, Villalba R, Cony M, Alberto M. Wood
A
productivity of Prosopis flexuosa D.C. woodlands in the central
Monte: Influence of population structure and tree-growth
habit. J Arid Environ 2011;75(1):7e13.
Dussart E, Lerner P, Peinetti R. Long term dynamics of 2
populations of Prosopis Caldenia Burkart. J Range Manag
1998;51(6):685e91.
Cano E, Estelrich H, Holgado H. Accio`n del fuego en los
estratos graminosos y arbustivos de un bosque de calden.
Rev la fac Agron (La Pampa, Arg.) 1985;1:81e96.
Boo R, Pelaez D, Bunting S, Mayor M, Ela O. Effect of fire on
woddy species in central semi-arid Argentina. J Arid Environ
1997;35(1):87e94.
Gorondi A. La reintroduccion del fuego en el ecosistema
bosque de calden. Rev la Fac Agron (La Pampa, Arg.)
1995;5:149e54.
Sipowicz A. Ecologa y manejo del fuego en el ecosistema del
caldenal. 1994. INTA Boletn divulg Tecnica: 51.
Llell J. Incendios de bosques en La Pampa. Rev la Fac Agron
(La Pampa, Arg.) 1990;5(1):29e33.
Steibel P, Troinani O. El genero Prosopis (Leguminosae) en la
Provincia de La Pampa. Rev la Fac Agron (La Pampa, Arg.)
1999;10(2):24e48.
Llorens E. Estrategias de manejo del caldenal para lograr una
maxima produccion sustentable. Nuestro Campo; 2000:13e7.
Ryan K. Predicting prescribed fire effects on trees in the
interior west. In: Alexander M, Bisgrove G, editors. The art
and Science of fire management. Alberta (Canada): Forestry
Canada Northwest region-Northern Forestry Centre; 1990.
Bravo S, Gimenez A, Moglia J. Efectos del fuego en la madera
de Prosopis alba Griseb. y Prosopis nigra (Griseb.) Hieron,
Mimosaceae. Bosque 2001;22(1):51e63.
Huang C, Marsh SE, McClaran MP, Archer SR. Postfire stand
structure in a semiarid savanna: cross-scale challenges
estimating biomass. Ecol Appl 2007;17(7):1899e910.
Midgley JJ. Why the worlds vegetation is not totally
dominated by resprouting plants; because resprouters are
shorter than reseeders. Ecography 1996;19(1):92e6.
Wirth C, Schumacher J. Generic biomass functions for
Norway spruce in Central Europeea meta-analysis approach
toward prediction and uncertainty estimation. Tree Physiol
2004;24(2):121e39.
Ketterings QM, Coe R, van Noordwjjk M, Ambagau Y,
Palm CA. Reducing uncertainty in the use of allometric
biomass equations for predicting above-ground tree biomass
in mixed secondary forests. For Ecol Manag
2001;146(1e3):199e209.
Ledesma M, Carranza CA, Balzarini M. Estimacion de la
biomasa foliar de Prosopis flexuosa mediante relaciones
alometricas. AgriScientia 2010;27(2):87e96.
Ruiz-Peinado R, Montero G, del Ro M. Biomass models to
estimate carbon stocks for hardwood tree species. For Syst
2012;21(1):42e52.
Mokany K, Raison J. Critical analysis of root:shoot ratios in
terrestrial biomass. Glob Change Biol 2006;12(1):84e96.
Cavelier J. Fine-root biomass and soil properties in a semideciduous tropical forest in Mexico. Plant Soil
1992;142(2):187e201.
Gower S. Relations between mineral nutrient availability and
fine root biomass in two Costa Rica tropical forest wet
forests: a hypothesis. Biotropica 1987;19(2):171e5.
Gerhardt K, Fredriksson D. Biomass allocation by broadleaf
mahogany seedlings, Swietenia microphylla (King), in
abandoned pastures and secondary dry forest in
Guanascaste, Costa Rica. Biotropica 1995;27(2):171e5.

Please cite this article in press as: Risio L, et al., Aboveground and belowground biomass allocation in native Prosopis caldenia
Burkart secondaries woodlands in the semi-arid Argentinean pampas, Biomass and Bioenergy (2014), http://dx.doi.org/10.1016/
j.biombioe.2014.03.038

12

b i o m a s s a n d b i o e n e r g y x x x ( 2 0 1 4 ) 1 e1 2

[63] Keyes M, Grier C. Above and belowground net production in

40-year-old Douglas-fir stands on low and high productivity
sites. Can J For Res 1981;11(3):599e605.
[64] Litton C, Ryan M, Tinker D, Knight D. Belowground and
aboveground biomass in young postfire lodgepole pine forest
of contrasting tree density. Can J For Res 2003;33(2):351e63.
[65] Bogino S, Jobbagy E. Climate and groundwater effects on the
establishment, growth and death of Prosopis caldenia trees in
the Pampas (Argentina). For Ecol Manag 2011;262(9):1766e74.
[66] Cannel MG. World forest biomass and primary production
data. London (England): Academic Press; 1982.

[67] Cairns M, Brown S, Helmer EH, Baumgardner G. Root

biomass allocation in the worlds upland forests. Oecologia
1997;111(1):1e11.
[68] Gaston G, Brown S, Lorenzini M, Singhi K. State and change
in carbon pools in the forests of tropical Africa. Glob Change
Biol 1998;4(1):97e114.
[69] Herrero C, Bravo F. Can we get an operational indicator of
forest carbon sequestration? A case study from two forest
regions in Spain. Ecol Indicat 2012;17:120e6.

Please cite this article in press as: Risio L, et al., Aboveground and belowground biomass allocation in native Prosopis caldenia
Burkart secondaries woodlands in the semi-arid Argentinean pampas, Biomass and Bioenergy (2014), http://dx.doi.org/10.1016/
j.biombioe.2014.03.038